<![CDATA[This research aimed to restructure the backend architecture of the XYZ application prototype using Microservice architecture. Load testing was conducted to compare the performance of the initial prototype backend and microservice architecture on response-time, throughput, and latency metrics. The restructuring method used was arranged in 4 stages with a total of 7 activities. In the first stage, system analysis was conducted on the XYZ application prototype. In the second stage, architecture decomposition, consisting of 3 activities, which were identifying system operations, identifying services using domain-driven design decomposition, and defining services and collaboration, was performed. In the third stage, database requirements analysis was performed on the microservice architecture that had been formed. In the fourth stage, the database design and microservice backend were implemented and tested using 3 different amounts of data, which were 56, 112, and 210, against 14 endpoints on both the prototype backend and the microservice backend. Based on the test results taken from the Apache JMeter Listener, it showed that the prototype backend showed superior performance in testing per endpoint, but in the overall test, the microservice backend showed better performance with a 2,4% faster response time, 1,8% higher throughput, and 2,4% lower latency. There is a pattern that shows the dominant microservice backend excels in tests with the last data, Data 210, in all metrics measured.]]>
